• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제15권10호
• /
• pp.3729-3749
• /
• 2021

At present, deep convolution network-based salient object detection (SOD) has achieved impressive performance. However, it is still a challenging problem to make full use of the multi-scale information of the extracted features and which appropriate feature fusion method is adopted to process feature mapping. In this paper, we propose a new adjacency auxiliary network (AANet) based on multi-scale feature fusion for SOD. Firstly, we design the parallel connection feature enhancement module (PFEM) for each layer of feature extraction, which improves the feature density by connecting different dilated convolution branches in parallel, and add channel attention flow to fully extract the context information of features. Then the adjacent layer features with close degree of abstraction but different characteristic properties are fused through the adjacent auxiliary module (AAM) to eliminate the ambiguity and noise of the features. Besides, in order to refine the features effectively to get more accurate object boundaries, we design adjacency decoder (AAM_D) based on adjacency auxiliary module (AAM), which concatenates the features of adjacent layers, extracts their spatial attention, and then combines them with the output of AAM. The outputs of AAM_D features with semantic information and spatial detail obtained from each feature are used as salient prediction maps for multi-level feature joint supervising. Experiment results on six benchmark SOD datasets demonstrate that the proposed method outperforms similar previous methods.

• 한국정보통신학회논문지
• /
• 제26권3호
• /
• pp.389-395
• /
• 2022

심층 학습 기술의 발전으로 인해 분류, 객체 검출, 분할과 같은 시각 정보를 이용한 심층 학습이 다양한 분야에서 활용되고 있다. 그 중 자율 주행은 시각 데이터를 잘 활용하는 대표적인 분야 중 하나이다. 본 논문에서는 도로 위의 사람과 운송수단 객체에 대한 개별적인 깊이 값을 예측하는 망을 제안한다. 제안하는 모델은 YOLOv3와 Monodepth를 기반으로 하며, 하드 파라미터 쉐어링을 이용한 인코더와 디코더를 통해 객체 검출과 깊이 추정을 동시에 수행한다. 또한 주의 집중 기법을 사용하여 객체 검출 및 깊이 추정의 정확도를 높이고자 하였다. 깊이 추정은 단안 이미지를 통해 이루어지며, 자가 학습 방법을 통해 학습을 수행하였다.

• International Journal of Contents
• /
• 제2권1호
• /
• pp.5-8
• /
• 2006

This paper describes a new mechanism of detecting object of interest from a noisy image, without using any a-priori knowledge about the target. It employs a parallel set of filters inspired upon biological findings of mammalian vision. In our proposed system, several basic features are extracted directly from original input visual stimuli, and these features are integrated based on their local competitive relations and statistical information. Through integration process, unnecessary features for detecting the target are spontaneously decreased, while useful features are enhanced. Experiments have been performed on a set of computer generated and real images corrupted with noise.

• 한국군사과학기술학회지
• /
• 제25권5호
• /
• pp.443-461
• /
• 2022

Target Detection in synthetic aperture radar(SAR) image is critical for military and national defense. In this paper, we propose YOLOv4-Attention architecture which adds attention modules to YOLOv4 backbone architecture to complement the feature extraction ability for SAR target detection with high accuracy. For training and testing our framework, we present new SAR embedding datasets based on MSTAR SAR public datasets which are about poor environments for target detection such as various clutter, crowded objects, various object size, close to buildings, and weakness of signal-to-clutter ratio. Experiments show that our Attention YOLOv4 architecture outperforms original YOLOv4 architecture in SAR image target detection tasks in poor environments for target detection.

• ETRI Journal
• /
• 제25권2호
• /
• pp.121-132
• /
• 2003

To enhance the productivity of software development and accelerate time to market, software developers have recently paid more attention to a component-based development (CBD) approach due to the benefits of component reuse. Among CBD processes, the identification of reusable components is a key but difficult process. Currently, component identification depends mainly on the intuition and experience of domain experts. In addition, there are few systematic methods or tools for component identification that enable domain experts to identify reusable components. This paper presents a systematic method and its tool called a component identifier that identifies software components by using object-oriented domain information, namely, use case models, domain object models, and sequence diagrams. To illustrate our method, we use the component identifier to identify candidates of reusable components from the object-oriented domain models of a banking system. The component identifier enables domain experts to easily identify reusable components by assisting and automating identification processes in an earlier development phase.

• 한국인터넷방송통신학회논문지
• /
• 제11권1호
• /
• pp.185-191
• /
• 2011

최근들어 영상보안 시스템에 관한 관심이 높아지고 있다. 영상으로부터 객체를 검출하고, 객체가 사람인지를 판별하며, 인식하는 기술이 다방면으로 활용되고 있다. 따라서 본 논문에서는 이러한 객체를 검색하기 위한 적응적인 방법을 제안하며, 이를 위하여 지역적 컨텍스트 기반의 얼굴 특징 검출 방법을 제안한다. 가보 번치를 이용하여 검출하는 이와 함께 베이지안 검출 방법을 이용한 특징점 보정에 따른 특징 검색 방법을 설명한다. 전체적인 시스템은 영상에서 오브젝트 영역을 검색하고, 지역적 컨텍스트 기반의 얼굴 검출, 특징 추출 방법을 적용하여 시스템의 성능을 높인다.

• 한국정보과학회:학술대회논문집
• /
• 한국정보과학회 2004년도 봄 학술발표논문집 Vol.31 No.1 (B)
• /
• pp.772-774
• /
• 2004

선택적 주의선택 알고리즘 중 색상과 밝기 지도를 만드는 부분을 실시간 처리에 적용하고, 저주파수가 주로 분포하는 영역과 물체의 움직임이 감지된 영역을 입력 영상에서 찾아내어 가중치별로 합산함으로서 실시간으로 선택적 주의를 줄 수 있는 시스템을 구현했다.

• 제어로봇시스템학회논문지
• /
• 제11권7호
• /
• pp.587-594
• /
• 2005

In this paper, a robust correlation-based video tracking is proposed to track a moving object in correlated image sequences. A correlation-based video tracking algorithm seeks to align the incoming target image with the reference target block image, but has critical problems, so called a false-peak problem and a drift phenomenon (correlator walk-off. The false-peak problem is generally caused by highly correlated background pixels with similar intensity of a moving target and the drift phenomenon occurs when tracking errors accumulate from frame to frame because of the nature of the correlation process. At first, the false-peaks problem for the ordinary correlation-based video tracking is investigated using a simple mathematical analysis. And, we will suggest a robust selective-attention correlation measure with a gradient preprocessor combined by a drift removal compensator to overcome the walk-off problem. The drift compensator adaptively controls the template block size according to the target size of interest. The robustness of the proposed method for practical application is demonstrated by simulating two real-image sequences.

• 대한원격탐사학회:학술대회논문집
• /
• 대한원격탐사학회 2002년도 Proceedings of International Symposium on Remote Sensing
• /
• pp.157-162
• /
• 2002

Recently, location based services, which make use of location information of moving objects, have obtained increasingly high attention. The moving objects are time-evolving spatial objects, that is, their locations are dynamically changed as time varies. Generally, GIS server stores and manages the spatial objects, of which locations are rarely changed. The traditional GIS server, however, has a difficulty to manage the moving objects, due to the fact of locations being frequently changed and the trajectory information (past locations of moving objects) being managed. In this paper, we have designed a moving object server, which stores and manages the locations in order to support various location based services. The moving object server is composed of a location acquisition component, a location storage component, and a location query component. The contribution of this paper is that we integrate the each work for location acquisition, storage, and query into a moving objects server.

• 반도체디스플레이기술학회지
• /
• 제19권2호
• /
• pp.60-67
• /
• 2020

Recently, artificial intelligence related technologies including machine learning are being applied to various fields, and the demand is also increasing. In particular, with the development of AR, VR, and MR technologies related to image processing, the utilization of computer vision based on deep learning has increased. The algorithms for object recognition and detection based on deep learning required for image processing are diversified and advanced. Accordingly, problems that were difficult to solve with the existing methodology were solved more simply and easily by using deep learning. This paper introduces various deep learning-based object recognition and extraction algorithms used to detect and recognize various objects in an image and analyzes the technologies that attract attention.